ÔÚ 2017-12-03 02:51:03£¬"Bruce Fischl" <fischl@nmr.mgh.harvard.edu> дµÀ£º
>I think you probably should since you don't want whatever you find to 
>only reflect differences in head size
>
>cheers
>Bruce
>On Sat, 2 Dec 2017, Ö£·ïÁ« wrote:
>
>> Hi Bruce,
>> 
>>     I think I get it. I have 27 males and 28 females and 26 males. It isn't necessary to corrected by eTIV. Is it right?
>> 
>> Thanks a lot !
>> Zheng
>> 
>> 
>> 
>> 
>> 
>> At 2017-12-02 11:45:10, "Bruce Fischl" <fischl@nmr.mgh.harvard.edu> wrote:
>> >Hi Zheng
>> >
>> >For volume, usually including eTIV reduces the variance and increases power 
>> >more than you lose for having one additional regressor. Not sure about 
>> >gender - how unbalanced is your sample? You might be ok with just eTIV
>> >Bruce
>> >
>> >
>> >On Sat, 2 Dec 2017, Ö£·ïÁ« 
>> >wrote:
>> >
>> >> Hi professor,
>> >> 
>> >>    I am using Freesurfer 6.0 and get the volume of subcortical structures. I want to get the relationship between age a
>> nd
>> >> volume. I only have one group, is it necessary to take gender and  eTIV as a covariate? If it is necessary, Which
>> >> statistical analysis should be used?
>> >>    I really appreciate your help. Thanks a lot.
>> >> 
>> >> Sincerely,
>> >> Zheng
>> >> 
>> >> 
>> >> 
>> >>
>> 
>> 
>> 
>>  
>> 
>> 
>>